BTP Resource Consumption Monitor

Description

Application to monitor and alert on commercial and technical consumption of BTP services. See also SAP Blog: Prototype: Keep Track of those BTP Credits.

Disclaimer: This tool is provided as-is and is not covered by SAP Support. This is not a replacement of the official billing documents you receive from SAP. The information provided by this tool is purely indicative.

Requirements

The following Subscriptions are required to deploy and use this application:

• SAP HANA Cloud (you can re-use an existing instance)
• SAP Work Zone (Standard edition is sufficient)
• SAP Business Application Studio

The following Entitlements need to be available to use this application:

• Alert Notification: standard
• Authorization and Trust Management Service: application
• Destination service: lite
• HTML5 Application Repository Service: app-host
• Job Scheduling Service: standard
• SAP HANA Schemas & HDI Continers: hdi-shared
• Usage Data Management Service: reporting-ga-admin
• Application Logging Service: standard (optional service)

Download and Installation

1. CF Application

In Business Application Studio, make sure to have a Development Space of kind Full Stack Cloud Application with the additional Development Tools for SAP Build Work Zone extension enabled.

Clone this repository in your environment and open the project.

Important - First change the following code lines:

• Configure your email address in the notifications configuration file mtaext_notifications.mtaext
• Configure your subaccount id in business app configuration files of both ui5 applications:
  1. btprcreport app: app.btprcreport.json
  2. managealerts app: app.managealerts.json

Option A. Deploy with Alert Notification configuration (existing configuration will be overwritten):

cd cf
npm install
mbt build
cf deploy ./mta_archives/btp-resource-consumption_1.0.0.mtar -e mtaext_notifications.mtaext

Option B. Deploy without changing any Alert Notification configuration:

cd cf
npm install
mbt build
cf deploy ./mta_archives/btp-resource-consumption_1.0.0.mtar

2. Work Zone Content

Step 1. Build

cd workzone
npm install
npm run build

This will create a /workzone/package.zip file which you can download to your local machine to use in step 2 below.

Step 2.Deploy the package

In the Work Zone Site Manager, open the Channel Manager and:

1. Synchronize your HTML5 Repository
2. Click on + New, Content Package and upload the generated /workzone/package.zip file, specifying btprc-srv as Runtime Destination
3. Navigate to the Site Directory and make sure you have a site with the view setting set to Spaces and Pages - New Experience, and which has the BTP Resource Consumption Role role assigned.

3. Role Assignments

In the BTP Cockpit, go to the Security settings of your subaccount and assign the below 2 Role Collections to your user:

• ~btprc.cpkg_access_role to access the Work Zone content (front-end)
• BTPResourceConsumption Viewer to access the CF Application (back-end)

Access the application

Open the Work Zone site and navigate to the BTP Credits page.

It is suggested to configure the application and modify the 'Forecasting Configuration' individually for each of the services to make sure the forecast is calculated correctly. Until this is set correctly, the forecasted costs might be unrealisticly high.

Run Locally

To run the application locally, you need to bind to the cloud service instances and run the application in hybrid mode.

cds bind -2 btprc-uas,btprc-db,btprc-dest
cds-ts watch --profile hybrid

Open your browser to http://localhost:4004 where you will find the following Web Applications:

• /btprcreport/webapp
• /managealerts/webapp

Tip: This will connect to the HANA Cloud database. In case you want to use a local sqlite database for testing, run cds deploy -2 sqlite, and remove the binding to btprc-db from the .cdsrc-private.json file.

Architecture

Initial Data and Recurring Jobs

When started for the first time, the application creates multiple jobs in the Job Scheduling Service:

• Default_UpdateDailyUsage: a recurring job to retrieve the consumption information for that day. In its default configuration, this job will run every 3 hours (*/3 UTC). Typically, the Usage Data Management API provides new data points around 6am UTC but this can vary.
• Default_UpdateMonthlyUsage: a recurring job to retrieve the consumption information from the past month. In its default configuration, this job will run every 1st day of the month, at 1:00 AM and 3:00 AM UTC. It runs twice, in case the first run is not successful.
• Default_UpdateHistoricUsage: a one-time job which runs after creation of the job. It is used to retrieve the consumption information of that day + the past months (default: from October 2023, see settings.ts). This makes sure that after deploying the application you immediately have some information in the dashboards before the scheduled jobs run.

The initial data that this last job retrieves can be refreshed/extended from the UI: in the application open the Data Management menu and click on Load historic data. Here you can specify how far back you want to retrieve the consumption information.

The other menu option (Load consumption for current month) is exactly the same as what the first job runs every 3 hours, so it is not needed to be used and is mainly intented for testing purposes.

You can freely edit/change the created jobs, or create other jobs. When the application (re)starts it will re-create the jobs if they don't exist already, but it won't overwrite them if they do exist.

Note: Jobs have a default expected completion time of 15 seconds. Depending on the query, the application and/or API can take longer to process a request. In that case, the job monitor will classify the run as failed even though it ran successfully.

Forecasting Configuration

Each commercial metric of a service has a Forecasting Configuration. The following settings are available:

• Excluded: the metric will not be forecasted/propagated. The current consumption for today is what is expected for the entire month. This should be used for 'stable services'.
  • examples: SAP Integration Suite tenants, SAP Business Application Studio users, ...
• Time Linear: the metric will be forecasted with a linear trend, based on time. The consumption will be divided by the current day in the month and multiplied by 30 to come to a full month's consumption. This should be used for services that have a steady consumption throughout the month.
  • examples: SAP HANA Cloud capacity units, Cloud Foundry GB memory, ...
• Time Degressive: the metric will be forecasted with a degressive/progressive trend, based on time. This is a variant of the linear approach, but you can give a degression factor to specify the decrease/increase in consumption through the month. This should be used for services that have a free allowance and charges only apply above this allowance (progressive), or for services that have a baseline consumption + some small uptake throughout the month (degressive)
  • Degression factor values:
    • < 1: the usage of the remaining days will be lower than the usage of the past days (degressive)
    • = 1: the usage of the remaining days will be similar to the usage of the past days (linear)
    • > 1: the usage of the remaining days will be higher than the usage of the past days (progressive)
  • examples: SAP Integration Suite transactions (has free allowance), Mobile Services users (most users' first access will be early in the month), ...

The forecasting settings can be all reset to the default value (see settings.ts) from the UI: menu Forecast Management: Revert all forecast settings to default.

Switching Global Accounts

The application can be connected to a different Global Account to monitor that consumption instead of the Global Account where the application is deployed in (default).

To do so:

• Manually create an instance of the Usage Data Management Service service of plan reporting-ga-admin in the other Global Account. Create a Service Key on that service instance, and copy its contents.
• In the Global Account where the application is deployed, create a new User Provided Service in which you paste the service key contents.
• Adapt the mta.yaml on lines 35, 36 and 176, 183 to swap the bound standard service instance for the user-provided instance.

Using Free Tier

For demo and test purposes it is possible to implement this solution using Free Tier services. Be aware there are restrictions when using the Free Tier service plans.

• Subscriptions:

  • SAP HANA Cloud: free tier plan is available, but note your database will be stopped every night causing jobs to fail. Restart your database in the mornings. See restrictions.
  • SAP Work Zone, standard edition: free tier plan is available and can be used. See restrictions.
  • SAP Business Application Studio: free tier plan is available and can be used. See restrictions.

• Resources:

  • Cloud Foundry runtime: free tier plan is available and can be used. Create your CF Organisation with the 'free' option before deploying the application.
  • SAP Alert Notification service: free tier plan is available and can be used. Switch your MTA to plan 'free' instead of 'standard' here.
  • SAP Application Logging service: there is no free plan, but this is an optional service, so you can just remove it from your MTA, both here and here.
  • SAP Job Scheduling service: no free plan available. For an indicative pricing, see here.

Known Issues

• Dynamic Tile: In case the dynamic tile of the Report application does not show your forecasted CPEA credit for this month (but shows 3 dots instead), you will need to manually create a system mapping. In the Work Zone Site Manager, navigate to the Settings menu and go to Alias Mapping. Add a new alias with the following settings: Aliases = sid(BTPRC.CPKG) and Runtime Destination = btprc-srv. If this mapping already exists and the error is still present, edit and save this mapping without making changes to re-trigger it.
• Cachebuster error 500: In case your applications do not open from the launchpad because of a failed call to /cachebusterTokens.json, you can correct this from the Work Zone Channel Manager. Use the Update content button of the HTML5 Apps entry 2 times (refresh the content twice consecutively), which will create the cache records in the background and then your apps should open correctly.

How to obtain support

Create an issue in this repository if you find a bug or have questions about the content.

For additional support, ask a question in SAP Community.

Contributing

If you wish to contribute code, offer fixes or improvements, please send a pull request. Due to legal reasons, contributors will be asked to accept a DCO when they create the first pull request to this project. This happens in an automated fashion during the submission process. SAP uses the standard DCO text of the Linux Foundation.

Code of Conduct

See Our Code of Conduct.

License

Copyright (c) 2024 SAP SE or an SAP affiliate company. All rights reserved. This project is licensed under the Apache Software License, version 2.0 except as noted otherwise in the LICENSE file.